A group of toxic chemicals results in the intra-axonal accumulation of neurofilaments. Studies of the pathogenesis of these neurofilbrillary aggregates have shown that the accumulations result from an impaired transport of the neurofilaments down the axon in the slow component (SCa) of axonal transport. A major issue is whether the impairment of transport represents an alteration of the transport mechanism, or whether there is an inherent alteration of the neurofilament. Evidence from a series of Gamma-diketones suggests that the neurofilament accumulations are the result of an alteration of the neurofilament, specifically covalent crosslinking of the neurofilament. It is hypothesized that in the Gamma-diketone series the neurofibrillary pathology results from cyclization of the Gamma-diketone with lysyl residues to form pyrroles, autoxidation of the unstable pyrrolyl residues to form reactive intermediates, and subsequent attack of the reactive pyrrole species by an adjacent residue to form a covalent bridge. This covalent crosslinking of neurofilament proteins is proposed to be a cumulative process that ultimately results in neurofilamentous tangles which preclude their normal transport down the axon and lead to their accumulation. These findings are pertinent to several human disorders in which neurofibrillary changes are secondary to an apparent alteration of the neurofilament. The paired helical filaments of Alzheimer's disease contain some neurofilament determinants as well as some specific determinants. Neurofilamentous aggregates also occur in ALS, giant axonal neuropathy, Lewy bodies, and Pick bodies. The relevance of the Gamma-disktone model to the human disorders is accentuated by the facts that paired helical filaments of Alzheimer's disease appear to be high molecular weight neurofilament polymers, and that there is a high molecular weight component of purified neurofilament preparations in normal adult animals. The basis of this proposal is to examine the mechanism of neurofilamentous aggregation under the premise that the neurofilament bundles in these disorders are the result of an inherent alteration of the neurofilament itself. The following approach is proposed: (1) to identify and quantify pyrrole formation following exposure to Gamma-diketones, (2) to identify the chemical structure of the pyrrole-mediated crossbridges, (3) to document the crosslinking of neurofilament proteins and other long-lived proteins in vivo and correlate this process with the intra-axonal neurofilament accumulation, (4) to isolate and characterize the high molecular weight species in normal neurofilament preparations, (5) to identify at the ultrastructural level which subunits of the neurofilament are involved in the toxic neurofilament tangles by immunogold methodology, and (6) to apply these immunogold techniques in the study of human disease in which neurofibrillary pathology is a major component.